Paramagnetic rim lesions in early multiple sclerosis: a 7 Tesla imaging study

Paramagnetic rim lesions (PRL) are considered biomarkers of chronic active lesions (CAL) in multiple sclerosis. The chronological relationship between CAL and disease severity in multiple sclerosis is not fully defined. In this cross-sectional, prospective study, we investigate the relationship between PRL and clinical/radiological measures of disease severity at the time of diagnosis of multiple sclerosis, clinically isolated syndrome, or radiologically isolated syndrome. Sixty-four treatment naïve people with newly diagnosed multiple sclerosis, clinically isolated syndrome, or radiologically isolated syndrome were consecutively enrolled. Each subject underwent 7.0 Tesla magnetic resonance imaging (MRI) and neurocognitive assessments. The MRI protocol included whole brain 3-dimensional acquisitions of: (i) magnetisation prepared 2 rapid gradient echoes, T₂-weighted fluid attenuated inversion recovery, and (ii) single echo gradient echo susceptibility sequences, from which susceptibility weighted imaging maps were derived. Differences in clinical and radiological metrics between people with and without PRL were assessed using a Welch's two-sample t-test (continuous variables) and Pearson's Chi-square test (categorical variables). Associations between clinical and radiological variables were measured using Pearson's correlation analyses. Two subjects were eliminated from the analyses for the following reasons: one subject was receiving a disease-modifying agent at the time of the 7.0 Tesla MRI, and the second had several tumefactive and confluent T₂-lesions which could have biased the results. PRL were detected in 29 of 62 (46.8%) people. PRL positive people had a larger T₂-lesion burden and lower normalized volumes of the thalamus and amygdala, as well as thinning of a few cortical regions (P ≤ 0.05). PRL positive individuals performed worse on the 9-hole Peg Test (non-dominant hand) and the Brief Visuospatial Memory Test-revised (total recall component) (P ≤ 0.05). Some of these differences were lost upon Bonferroni correction. We demonstrate that as early as the time of diagnosis, the presence of PRL is associated with a few clinical and radiological measures of disease severity. Thus, PRL can be considered qualitative indicators of a more aggressive disease presentation.